28 like AI, machine learning, neural network , and Internet of things (IOT) are not sufficient (by themselves) to impart novelty to a patent claim. (III) Statutory subject matter and (IV) enablement are more difficult to understand but no less worthy of an inventor’s attention when preparing a patent application. Statutory subject matter is most easily explained as how well the claimed invention fits into one of the three categories we are considering (device, system, and method). Looking at statutory subject matter in the context of AI inventions, the question we need to ask ourselves is “What’s special about the hardware?” Let’s consider a first invention where a system analyzes data from sensors in a field using an AI algorithm. The inventor has studied the relevant prior publications in depth and is convinced that the AI algorithm is novel and inventive. Is that enough to get a patent granted on the claimed system? The answer is maybe. It depends on what the system does with the data generated by the AI algorithm. If the system only presents the compiled data generated by the AI algorithm to the user in a way that is easy to comprehend (e.g. a heat map), the claimed invention is probably not patentable. The reason is that the AI algorithm itself is excluded from patentability as a matter of law, and generation of a heat map (in 2024) is relatively simple data manipulation which the Examiner will dismiss as “insignificant post solution activity”. On the other hand, if the system uses the compiled data generated by the AI algorithm to do something, the claimed invention may very well be patentable. The reason is that the AI algorithm is now functioning to exert a real-world effect. This integrates the AI algorithm into the system in a way that contributes to the utility of the system, independent of human intervention. This distances the claimed system from a “general-purpose computer”, although the AI algorithm may be running on a general-purpose computer. Of course, the best case for the inventor is if the system includes a new hardware component (e.g. a new sensor type or a new type of switch) in addition to the inventive AI algorithm. These considerations are worthy of an inventor’s time because patent examiners (especially in the US) are anxious to reject an application as being directed towards non-statutory subject matter. If you wait until you see the
RkJQdWJsaXNoZXIy MjgzNzA=